https://github.com/tasninanika/convolutional_neural_network-cnn

A deep learning project to classify handwritten digits (0–9) using a Convolutional Neural Network (CNN). This project builds a CNN model with TensorFlow/Keras to classify digits with high accuracy.
https://github.com/tasninanika/convolutional_neural_network-cnn

keras tensorflow

Last synced: 2 months ago
JSON representation

A deep learning project to classify handwritten digits (0–9) using a Convolutional Neural Network (CNN). This project builds a CNN model with TensorFlow/Keras to classify digits with high accuracy.

• Host: GitHub
• URL: https://github.com/tasninanika/convolutional_neural_network-cnn
• Owner: tasninanika
• Created: 2025-08-26T03:36:48.000Z (11 months ago)
• Default Branch: main
• Last Pushed: 2025-08-26T03:53:41.000Z (11 months ago)
• Last Synced: 2025-08-26T05:36:54.363Z (10 months ago)
• Topics: keras, tensorflow
• Language: Jupyter Notebook
• Homepage:
• Size: 141 KB
• Stars: 1
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# 🔢 MNIST Digit Classification using CNN

A deep learning project to classify **handwritten digits (0–9)** using a **Convolutional Neural Network (CNN)**. The model is trained on the **MNIST dataset**.

---

## 📌 Project Overview

The **MNIST dataset** is a benchmark dataset in machine learning and computer vision. It consists of grayscale images of handwritten digits.

This project builds a **CNN model with TensorFlow/Keras** to classify digits with high accuracy.

---

## 📂 Dataset

* **Source:** [MNIST Dataset](http://yann.lecun.com/exdb/mnist/) (also available via `keras.datasets.mnist`)

* **Training Samples:** 60,000

* **Test Samples:** 10,000

* **Image Size:** 28 × 28 pixels (grayscale)

* **Classes:** 10 (digits 0–9)

---

## ⚙️ Tech Stack

![Python](https://img.shields.io/badge/Python-3776AB?style=for-the-badge\&logo=python\&logoColor=white)

![NumPy](https://img.shields.io/badge/Numpy-013243?style=for-the-badge\&logo=numpy\&logoColor=white)

![Matplotlib](https://img.shields.io/badge/Matplotlib-11557c?style=for-the-badge\&logo=plotly\&logoColor=white)

![TensorFlow](https://img.shields.io/badge/TensorFlow-FF6F00?style=for-the-badge\&logo=tensorflow\&logoColor=white)

![Keras](https://img.shields.io/badge/Keras-D00000?style=for-the-badge\&logo=keras\&logoColor=white)

---

## 🛠️ Project Workflow

### 🔹 Data Preprocessing

* Loaded MNIST dataset from `keras.datasets`

* Normalized pixel values (0–255 → 0–1)

* Reshaped data for CNN input (`28x28x1`)

* One-hot encoded target labels

### 🔹 CNN Architecture

* **Conv2D + ReLU + MaxPooling** layers

* **Flatten** layer

* **Dense (Fully Connected)** layers

* **Output layer** with Softmax activation

### 🔹 Model Training

* **Optimizer:** Adam

* **Loss Function:** Categorical Crossentropy

* **Metrics:** Accuracy

* Trained for multiple epochs with batch size tuning

### 🔹 Evaluation

* Achieved **\~99% training accuracy**

* Achieved **\~98% test accuracy**

* Visualized predictions with sample test images

---

## 📊 Results

✔️ High accuracy (>98%) on test data

✔️ Robust CNN architecture with minimal overfitting

✔️ Correctly classifies handwritten digits